OpenAI details how Legora builds its agentic operating system: from 20% of legal time on NDAs to under 1%
The case published by OpenAI quantifies the impact: 95% less time on NDAs, contracts 90% faster and 9.5 hours a week per lawyer.
OpenAI has published a detailed case on how Legora, the Swedish startup founded in Stockholm in 2023, is building what its CEO and co-founder Max Junestrand calls the agentic operating system for legal work: a platform running on OpenAI models that, according to the figures provided, cuts by more than 95 per cent the time the company's own legal team spends reviewing NDAs.
The numbers of the case
OpenAI's document quantifies the impact in three metrics. Legora's legal team went from spending around 20 per cent of its capacity reviewing non-disclosure agreements to under 1 per cent, thanks to a playbook built on its own historical agreements that gives incoming NDAs a first pass and flags clauses outside acceptable parameters. A Fortune 100 customer cut contract turnaround time by 90 per cent, saved more than 20 million dollars on external litigation and diligence, and returned 9.5 hours a week to each lawyer for higher-value work.
AI-native inside and out
The case highlights that Legora also uses AI internally: ChatGPT Work, Codex and coding agents are part of daily operations. Its CTO, Jacob Lauritzen, explains that teams can explore twenty solutions in parallel and pick the best one within days, and that incident diagnosis has gone from hours to about five minutes searching millions of lines of logs.
General Counsel Emma Kadri Bergström adds the scaling dimension: in the past year the company has gone from three offices to eighteen, closed three financing rounds and completed five acquisitions, growth the legal team has had to sustain. Her conclusion is blunt: AI does not just make you faster, it also increases the quality of the analysis.
From fast review to broad analysis
Alex Fortescue-Webb, Global Head of Legal Engineering, stresses breadth: where a traditional review forced teams to concentrate on the most valuable contracts, the Tabular Review workflow analyses thousands of contracts across multiple questions in minutes, so risks previously outside the reviewed subset are now surfaced for a lawyer's judgment.
The case also offers adoption lessons transferable to any law firm: when more than 75 per cent of leaders actively use the tool, median usage across the organisation triples; start with the problem, not the technology; and design for trust, with evaluation and transparency from the start.
The horizon: long-running agents and real-time collaboration
The case also anticipates the roadmap. Lauritzen describes two advances in parallel: agents capable of tackling long-duration tasks with greater context and autonomy, such as due diligence over millions of pages run overnight with the analysis ready by morning, and ever-faster models that allow real-time collaboration between the lawyer and the AI on a contract, an analysis or a piece of advice. Product teams review the new possibilities with each model leap, in cycles the CTO himself puts at two weeks.
Legora also offers a plugin for ChatGPT that allows its capabilities to be used within ChatGPT conversations, an integration that brings the platform closer to the workflows lawyers already use. Junestrand sums up the ambition with a phrase that defines the sector's moment: it is, he says, the best time in history to be a lawyer.
Source: OpenAI (28 October 2026).
Related links
Fuentes
- openai.com. Accessed 2 November 2026.
The specific data in this article come from the official sources cited.
